Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751368Ab0GUMxP (ORCPT ); Wed, 21 Jul 2010 08:53:15 -0400 Received: from cantor2.suse.de ([195.135.220.15]:60037 "EHLO mx2.suse.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750838Ab0GUMxO (ORCPT ); Wed, 21 Jul 2010 08:53:14 -0400 Date: Wed, 21 Jul 2010 14:53:12 +0200 Message-ID: From: Takashi Iwai To: Wei Ni Cc: Wu Fengguang , "'Pavel Hofman'" , "'alsa-devel'" , "'linux-kernel'" , "'akpm'" Subject: Re: Support MCP89 hdmi audio In-Reply-To: <6B4D417B830BC44B8026029FD256F7F1C29D065C70@HKMAIL01.nvidia.com> References: <6B4D417B830BC44B8026029FD256F7F1C29D065C70@HKMAIL01.nvidia.com> User-Agent: Wanderlust/2.15.6 (Almost Unreal) SEMI/1.14.6 (Maruoka) FLIM/1.14.9 (=?UTF-8?B?R29qxY0=?=) APEL/10.7 Emacs/23.1 (x86_64-suse-linux-gnu) MULE/6.0 (HANACHIRUSATO) MIME-Version: 1.0 (generated by SEMI 1.14.6 - "Maruoka") Content-Type: text/plain; charset=US-ASCII Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 4363 Lines: 126 At Wed, 21 Jul 2010 19:45:09 +0800, Wei Ni wrote: > > Hi, all > I'm an Nvidia engineer, I maintain the nvidia linux hdmi audio driver. > > An end-user report a problem with his TV attached to an MCP89. > ALSA won't send audio to it, because > PD=0: > HDMI hot plug event: Pin=5 Presence_Detect=1 ELD_Valid=0 > HDMI hot plug event: Pin=5 Presence_Detect=0 ELD_Valid=1 > > As you know, in linux driver patch_hdmi.c, it use (pind && eldv) > in unsolicited events. So the driver can't detect the monitor. > > >From the HAD specification > (http://download.intel.com/standards/hdaudio/pdf/HDA036.pdf): > > Original text for Unsolicited Response PD bit: > > Presence Detect: When this bit is set, sense measurement has > changed on the pin widget and software can use the pin sense > control verb to determine the current pin sense data state. > For analog pin widgets, this UR means that Presence Detect or > Impedance has changed on the pin widget. > For digital pin widgets, including HDMI pin widgets, this means > that presence detect (and optionally ELD valid bit) has > **changed**. [my emphasis] > > Updated text for Unsolicted Response PD bit: > > Presence Detect: This bit reflects the **present** [my > emphasis] state of the Pin Sense Presence Detect bit when the > unsolicited response is triggered. Software can optionally use > the pin sense control verb to determine the latest pin sense > data state. This bit implementation is only required for > digital display pin widget. Non digital display pin widget is > optional to implement this bit. > > For our MCP89, the values indicates the change of status of PD, not the actual value of PD, > which follows older description of UR for PD bit. > > So could we add a workaround for MCP89? We could use Pin Sense to determine again in > unsolicited events. If it return 0xc0000000, it means PD=1 and ELVD=1. > Or do you have any suggestions? Yes. The patch below does similar thing. Could you give it a try? thanks, Takashi --- diff --git a/sound/pci/hda/patch_hdmi.c b/sound/pci/hda/patch_hdmi.c index 86067ee..2fc5396 100644 --- a/sound/pci/hda/patch_hdmi.c +++ b/sound/pci/hda/patch_hdmi.c @@ -52,6 +52,10 @@ struct hdmi_spec { */ struct hda_multi_out multiout; unsigned int codec_type; + + /* misc flags */ + /* PD bit indicates only the update, not the current state */ + unsigned int old_pin_detect:1; }; @@ -616,6 +620,9 @@ static void hdmi_setup_audio_infoframe(struct hda_codec *codec, hda_nid_t nid, * Unsolicited events */ +static void hdmi_present_sense(struct hda_codec *codec, hda_nid_t pin_nid, + struct hdmi_eld *eld); + static void hdmi_intrinsic_event(struct hda_codec *codec, unsigned int res) { struct hdmi_spec *spec = codec->spec; @@ -632,6 +639,12 @@ static void hdmi_intrinsic_event(struct hda_codec *codec, unsigned int res) if (index < 0) return; + if (spec->old_pin_detect) { + if (pind) + hdmi_present_sense(codec, tag, &spec->sink_eld[index]); + pind = spec->sink_eld[index].monitor_present; + } + spec->sink_eld[index].monitor_present = pind; spec->sink_eld[index].eld_valid = eldv; diff --git a/sound/pci/hda/patch_nvhdmi.c b/sound/pci/hda/patch_nvhdmi.c index 3c10c0b..b0652ac 100644 --- a/sound/pci/hda/patch_nvhdmi.c +++ b/sound/pci/hda/patch_nvhdmi.c @@ -478,6 +478,7 @@ static int patch_nvhdmi_8ch_89(struct hda_codec *codec) codec->spec = spec; spec->codec_type = HDA_CODEC_NVIDIA_MCP89; + spec->old_pin_detect = 1; if (hdmi_parse_codec(codec) < 0) { codec->spec = NULL; @@ -508,6 +509,7 @@ static int patch_nvhdmi_8ch_7x(struct hda_codec *codec) spec->multiout.max_channels = 8; spec->multiout.dig_out_nid = nvhdmi_master_con_nid_7x; spec->codec_type = HDA_CODEC_NVIDIA_MCP7X; + spec->old_pin_detect = 1; codec->patch_ops = nvhdmi_patch_ops_8ch_7x; @@ -528,6 +530,7 @@ static int patch_nvhdmi_2ch(struct hda_codec *codec) spec->multiout.max_channels = 2; spec->multiout.dig_out_nid = nvhdmi_master_con_nid_7x; spec->codec_type = HDA_CODEC_NVIDIA_MCP7X; + spec->old_pin_detect = 1; codec->patch_ops = nvhdmi_patch_ops_2ch; -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/